Throughout this application, various references are referred to within parentheses. Disclosures of these publications in their entireties are hereby incorporated by reference into this application to more fully describe the state of the art to which this invention pertains. Full bibliographic citations for these references may be found at the end of this application, preceding the claims.
The carcinogenic process involves a series of sequential changes in the phenotype of a cell resulting in the acquisition of new properties or a further elaboration of transformation-associated traits by the evolving tumor cell (1-4). Although extensively studied, the precise genetic mechanisms underlying tumor cell progression during the development of most human cancers remain enigmas. Possible factors contributing to transformation progression, include: activation of cellular genes that promote the cancer cell phenotype, i.e., oncogenes; activation or modification of genes that regulate genomic stability, i.e., DNA repair genes; loss or inactivation of cellular genes that function as inhibitors of the cancer cell phenotype, i.e. tumor suppressor genes; and/or combinations of these genetic changes in the same tumor cell (1-6). A useful model system for defining the genetic and biochemical changes mediating tumor progression is the type 5 adenovirus (Ad5)/early passage rat embryo (RE) cell culture system (1,7-14). Transformation of secondary RE cells by Ad5 is often a sequential process resulting in the acquisition of and further elaboration of specific phenotypes by the transformed cell (7-10). Progression in the Ad5-transformation model is characterized by the development of enhanced anchorage-independence and tumorigenic potential (as indicated by a reduced latency time for tumor formation in nude mice) by progressed cells (1,10). The progression phenotype in Ad5-transformed RE cells can be induced by selection for growth in agar or tumor formation in nude mice (7-10), referred to as spontaneous-progression, by transfection with oncogenes (13), such as Ha-ras, v-src, v-raf or the E6/E7 region of human papillomavirus type (HPV)-18, referred to as oncogene-mediated progression, or by transfection with specific signal transducing genes (14), such as protein kinase C, referred to as growth factor-related, gene-induced progression.
Progression, induced spontaneously or after gene transfer, is a stable cellular trait that remains undiminished in Ad5-transformed RE cells even after extensive passage (&gt;100) in monolayer culture (13). However, a single-treatment with the demethylating agent 5-azacytidine (AZA) results in a stable reversion in transformation progression in &gt;95% of cellular clones (10,13,14). The progression phenotype is also suppressed in somatic cell hybrids formed between normal or unprogressed transformed cells and progressed cells (11-13). These findings suggest that progression may result from the activation of specific progression-promoting genes or the selective inhibition of progression-suppressing genes, or possibly a combination of both processes.
The final stage in tumor progression is acquisition by transformed cells of the ability to invade local tissue, survive in the circulation and recolonize in a new area of the body, i.e., metastasis (15-17). Transfection of a Ha-ras oncogene into cloned rat embryo fibroblast (CREF) cells (18) results in morphological transformation, anchorage-independence and acquisition of tumorigenic and metastatic potential (19-21). Ha-ras-transformed CREF cells exhibit major changes in the transcription and steady-state levels of genes involved in suppression and induction of oncogenesis (21,22). Simultaneous overexpression of the Ha-ras suppressor gene Krev-1 in Ha-ras-transformed CREF cells results in morphological reversion, suppression of agar growth capacity and a delay in in vivo oncogenesis (21). Reversion of transformation in Ha-ras+Krev-1 transformed CREF cells correlates with a return in the transcriptional and steady-state mRNA profile to that of untransformed CREF cells (21,22). Following long latency times, Ha-ras+Krev-1 transformed CREF cells form both tumors and metastases in athymic nude mice (21). The patterns of gene expression changes observed during progression, progression suppression and escape from progression suppression supports the concept of "transcriptional switching" as a major component of Ha-ras-induced transformation (21,22).
To identify potential progression inducing genes with elevated expression in progressed versus unprogressed Ad5-transformed cells we used subtraction hybridization (13,23). This approach resulted in the cloning of PEG-3 that is expressed at elevated levels in progressed cells (spontaneous, oncogene-induced and growth factor-related, gene-induced) than in unprogressed cells (parental Ad5-transformed, AZA-suppressed, and suppressed hybrids). Transfection of PEG-3 into unprogressed parental Ad5-transformed cells induces the progression phenotype, without significantly altering colony formation in monolayer culture or affecting cell growth. PEG-3 expression is also elevated following DNA damage and oncogenic transformation of CREF cells by various oncogenes. Sequence analysis indicates that PEG-3 has 73 and 68% nucleotide (nt) and 59 and 72% amino acid (aa) similarities, respectively, with the gadd34 and MyD116 gene. However, unlike gadd34 and MyD116 that encode proteins of .about.65 and .about.72 kDa, respectively, PEG-3 encodes a protein of .about.50 kDa with only .about.28 and .about.40% aa similarities to gadd34 and MyD116, respectively, in its carboxyl terminus. These results indicate that PEG-3 represents a new member of the gadd34/MyD116 gene family with both similar and distinct properties. Unlike gadd34 and MyD116, which dramatically suppress colony formation (24), PEG-3 only modestly alters colony formation following transfection, i.e., .ltoreq.20% reduction in colony formation in comparison with vector transfected cells. Moreover, a direct correlation only exists between expression of PEG-3, and not gadd34 or Myd116, and the progression phenotype in transformed rodent cells. These findings provide evidence for a potential link between constitutive induction of a stress response, characteristic of DNA damage, and induction of cancer progression.